NumPy allows the creation of n-dimensional arrays, which is where the name of the data type, numpy.ndarray, comes from. It handles many sophisticated scientific and matrix operations. It provides many linear algebra and random number functionalities.
NumPy lies at the core of many calculations that computationally enable Matplotlib and many other Python packages. It is therefore a dependency for many common packages and often comes along with Python distributions. For instance, it provides the fundamental data structure for SciPy, a package that handles statistical calculations useful in science and many other areas.
To import NumPy, input this:
import numpy as np
To create a NumPy array from lists, use the following:
x = np.array([2,3,1,0])
You can also create non-integral arithmetic series with NumPy by using np.linspace(start,stop,number).
See the following example:
In [1]: np.linspace(3,5,20) Out[1]: array([ 3. , 3.10526316, 3.21052632, 3.31578947, 3.42105263, 3.52631579, 3.63157895, 3.73684211, 3.84210526, 3.94736842, 4.05263158, 4.15789474, 4.26315789, 4.36842105, 4.47368421, 4.57894737, 4.68421053, 4.78947368, 4.89473684, 5. ])
Matrix operations can be applied across NumPy arrays. Here is an example of multiplying two arrays:
In [2]: a = np.array([1, 2, 1])
In [3]: b = np.array([2, 3, 8])
In [4]: a*b
Out[4]: array([2, 6, 8])